Re: Preliminary EARL report

Hello!

2012/3/15 Gregg Kellogg <gregg@kellogg-assoc.com>:
> It runs quite fast against many of the host languages, but doesn't seem to return for HTML4 or XHTML1; any idea why that might be? It seems to return 503 "Service Unavailable".

Yes, at the time I only used the standard XML parsing support in Java.
Ad the XHTML failed because it tried to fetch the XHTML DTD. I have
now included NekoHTML to support HTML, and turned off the DTD
fetching. So I can happily report that clj-rdfa passes all tests for
all markup flavours! Well, except for #0198 in XHTML5...

> Regarding Test 0198, mine returns the following:
>
> @base <http://rdfa.info/test-suite/test-cases/rdfa1.1/xhtml1/0198.xhtml> .
> @prefix foaf: <http://xmlns.com/foaf/0.1/> .
> @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
>
> <http://www.example.org/me#mark> a foaf:Person;
>   foaf:firstName "Mark";
>   foaf:name "<span xmlns=\"http://www.w3.org/1999/xhtml\" xmlns:foaf=\"http://xmlns.com/foaf/0.1/\" xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\" property=\"foaf:firstName\">Mark</span> <span xmlns=\"http://www.w3.org/1999/xhtml\" xmlns:foaf=\"http://xmlns.com/foaf/0.1/\" xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\" property=\"foaf:surname\">Birbeck</span>"^^rdf:XMLLiteral;
>   foaf:surname "Birbeck" .
>
> For XHTML1 and XHTML5, This does include an @xmlns, and it passes the test; what is it that you think should be different?

In the test suite data, the markup includes the expected @xmlns when
requested as </test-suite/test-cases/rdfa1.1/xhtml1/0198.xhtml>, but
it is missing in </test-suite/test-cases/rdfa1.1/xhtml5/0198.xhtml>.
Implementation-wise, it seems that if the template
"lib/views/test_case.xhtml.haml" was used, it would add this. But in
core.rb, the "when 'xhtml'" case doesn't for xhtml5. Isn't that a bug?

Best regards,
Niklas


> Gregg
>
> On Mar 14, 2012, at 4:56 PM, Niklas Lindström wrote:
>
>> Hello!
>>
>> 2012/3/14 Gregg Kellogg <gregg@kellogg-assoc.com>:
>>> On Mar 14, 2012, at 8:28 AM, Niklas Lindström wrote:
>>>
>>>> My problems are related to multilines, and these tests pass if I run
>>>> them manually using ARQ [2].
>>>
>>> Do you think the SPARQL is wrong, or that the processor isn't interpreting this properly? Send me your result output and I'll check it out.
>>
>> The culprit was the mime-type I sent: "text/plain; charset=utf-8".
>> While "text/plain" is for ntriples, I actually produce turtle-style
>> UTF-8 literals using triple quotes when encountering newlines. So I
>> corrected that to "text/turtle" and all those tests pass as well! I'm
>> down to only one remaining now:
>>
>>>> Other than that, my implementation only
>>>> fails on Test 0198 [3], which is missing an @xmlns. That may be a bug
>>>> in the test though?
>>>
>>> Yes, I think it is. I'll look at that.
>>
>> If xmlns="http://www.w3.org/1999/xhtml" is added to "tests/0198.txt",
>> clj-rdfa will pass all the tests. (Yay!)
>>
>> I have deployed my implementation to Heroku (the free option, so it
>> might take a while to start up at times). It might be good to set up a
>> proper domain for it before it's ready to be added to the harness
>> though. But you can try it out now by using
>> "http://clj-rdfa.herokuapp.com/extract.ttl?url=" as the value for
>> 'other' processor.
>>
>> (I'm also improving the interface for it so it can be seamlessly used
>> in plain Java, and thus adapted to RDF libraries such as Jena and
>> Sesame. When that's done it'll be a working RDFa 1.1 option for Java
>> in general.)
>>
>> Best regards,
>> Niklas
>

Received on Saturday, 17 March 2012 18:42:00 UTC